Google Is the Only Search Engine That Works on Reddit Now Thanks to AI Deal

Kinda long, so I'm putting it in spoilers. This applies to Nginx, but you can probably adapt it to other reverse proxies.

1. Create a file to hold the mappings and store it somewhere you can include it from your other configs. I named mine map-bot-user-agents.conf

Here, I'm doing a regex comparison against the user agent ($http_user_agent) and mapping it to either a 0 (default/false) or 1 (true) and storing that value in the variable $ua_disallowed. The run-on string at the bottom was inherited from another admin I work with, and I never bothered to split it out.

# Map bot user agents
map $http_user_agent $ua_disallowed {
    default 		0;
    "~CCBot"		1;
    "~ClaudeBot"	1;
    "~VelenPublicWebCrawler"	1;
    "~WellKnownBot"	1;
    "~Synapse (bot; +https://github.com/matrix-org/synapse)" 1;
    "~python-requests"	1;
    "~bitdiscovery"	1;
    "~bingbot"		1;
    "~SemrushBot" 	1;
    "~Bytespider" 	1;
    "~AhrefsBot" 	1;
    "~AwarioBot"	1;
    "~GPTBot" 		1;
    "~DotBot"	 	1;
    "~ImagesiftBot"	1;
    "~Amazonbot"	1;
    "~GuzzleHttp" 	1;
    "~DataForSeoBot" 	1;
    "~StractBot"	1;
    "~Googlebot"	1;
    "~Barkrowler"	1;
    "~SeznamBot"	1;
    "~FriendlyCrawler"	1;
    "~facebookexternalhit" 1;
    "~*(?i)(80legs|360Spider|Aboundex|Abonti|Acunetix|^AIBOT|^Alexibot|Alligator|AllSubmitter|Apexoo|^asterias|^attach|^BackDoorBot|^BackStreet|^BackWeb|Badass|Bandit|Baid|Baiduspider|^BatchFTP|^Bigfoot|^Black.Hole|^BlackWidow|BlackWidow|^BlowFish|Blow|^BotALot|Buddy|^BuiltBotTough|
^Bullseye|^BunnySlippers|BBBike|^Cegbfeieh|^CheeseBot|^CherryPicker|^ChinaClaw|^Cogentbot|CPython|Collector|cognitiveseo|Copier|^CopyRightCheck|^cosmos|^Crescent|CSHttp|^Custo|^Demon|^Devil|^DISCo|^DIIbot|discobot|^DittoSpyder|Download.Demon|Download.Devil|Download.Wonder|^dragonfl
y|^Drip|^eCatch|^EasyDL|^ebingbong|^EirGrabber|^EmailCollector|^EmailSiphon|^EmailWolf|^EroCrawler|^Exabot|^Express|Extractor|^EyeNetIE|FHscan|^FHscan|^flunky|^Foobot|^FrontPage|GalaxyBot|^gotit|Grabber|^GrabNet|^Grafula|^Harvest|^HEADMasterSEO|^hloader|^HMView|^HTTrack|httrack|HTT
rack|htmlparser|^humanlinks|^IlseBot|Image.Stripper|Image.Sucker|imagefetch|^InfoNaviRobot|^InfoTekies|^Intelliseek|^InterGET|^Iria|^Jakarta|^JennyBot|^JetCar|JikeSpider|^JOC|^JustView|^Jyxobot|^Kenjin.Spider|^Keyword.Density|libwww|^larbin|LeechFTP|LeechGet|^LexiBot|^lftp|^libWeb|
^likse|^LinkextractorPro|^LinkScan|^LNSpiderguy|^LinkWalker|msnbot|MSIECrawler|MJ12bot|MegaIndex|^Magnet|^Mag-Net|^MarkWatch|Mass.Downloader|masscan|^Mata.Hari|^Memo|^MIIxpc|^NAMEPROTECT|^Navroad|^NearSite|^NetAnts|^Netcraft|^NetMechanic|^NetSpider|^NetZIP|^NextGenSearchBot|^NICErs
PRO|^niki-bot|^NimbleCrawler|^Nimbostratus-Bot|^Ninja|^Nmap|nmap|^NPbot|Offline.Explorer|Offline.Navigator|OpenLinkProfiler|^Octopus|^Openfind|^OutfoxBot|Pixray|probethenet|proximic|^PageGrabber|^pavuk|^pcBrowser|^Pockey|^ProPowerBot|^ProWebWalker|^psbot|^Pump|python-requests\/|^Qu
eryN.Metasearch|^RealDownload|Reaper|^Reaper|^Ripper|Ripper|Recorder|^ReGet|^RepoMonkey|^RMA|scanbot|SEOkicks-Robot|seoscanners|^Stripper|^Sucker|Siphon|Siteimprove|^SiteSnagger|SiteSucker|^SlySearch|^SmartDownload|^Snake|^Snapbot|^Snoopy|Sosospider|^sogou|spbot|^SpaceBison|^spanne
r|^SpankBot|Spinn4r|^Sqworm|Sqworm|Stripper|Sucker|^SuperBot|SuperHTTP|^SuperHTTP|^Surfbot|^suzuran|^Szukacz|^tAkeOut|^Teleport|^Telesoft|^TurnitinBot|^The.Intraformant|^TheNomad|^TightTwatBot|^Titan|^True_Robot|^turingos|^TurnitinBot|^URLy.Warning|^Vacuum|^VCI|VidibleScraper|^Void
EYE|^WebAuto|^WebBandit|^WebCopier|^WebEnhancer|^WebFetch|^Web.Image.Collector|^WebLeacher|^WebmasterWorldForumBot|WebPix|^WebReaper|^WebSauger|Website.eXtractor|^Webster|WebShag|^WebStripper|WebSucker|^WebWhacker|^WebZIP|Whack|Whacker|^Widow|Widow|WinHTTrack|^WISENutbot|WWWOFFLE|^
WWWOFFLE|^WWW-Collector-E|^Xaldon|^Xenu|^Zade|^Zeus|ZmEu|^Zyborg|SemrushBot|^WebFuck|^MJ12bot|^majestic12|^WallpapersHD)" 1;

}

Once you have a mapping file setup, you'll need to do something with it. This applies at the virtual host level and should go inside the server block of your configs (except the include for the mapping config.).

This assumes your configs are in conf.d/ and are included from nginx.conf.

The map-bot-user-agents.conf is included above the server block (since it's an http level config item) and inside server, we look at the $ua_disallowedvalue where 0=false and 1=true (the values are set in the map).

You could also do the mapping in the base nginx.conf since it doesn't do anything on its own.

If the $ua_disallowed value is 1 (true), we immediately return an HTTP 444. The 444 status code is an Nginx thing, but it basically closes the connection immediately and wastes no further time/energy processing the request. You could, optionally, redirect somewhere, return a different status code, or return some pre-rendered LLM-generated gibberish if your bot list is configured just for AI crawlers (because I'm a jerk like that lol).

include conf.d/includes/map-bot-user-agents.conf;

server {
  server_name  example.com;
  ...
  # Deny disallowed user agents
  if ($ua_disallowed) { 
    return 444;
  }
 
  location / {
    ...
  }

}

How much do those bots consume your bandwidth?

Pretty negligible per bot per request, but I'm not here to feed them. They also travel in packs, so the bandwidth does multiply. It also costs me money when I exceed my monthly bandwidth quota. I've blocked them for so long, I no longer have data I can tally to get an aggregate total (I only keep 90 days). SemrushBot alone, before I blocked it, was averaging about 15 GB a month. That one is fairly aggressive, though. Imagesift Bot, which pulls down any images it can find, would also use quite a bit, I imagine, if it were allowed.

With Lemmy, especially earlier versions, the queries were a lot more expensive, and bots hitting endpoints that triggered a heavy query (such as a post with a lot of comments) would put unwanted load on my DB server. That's when I started blocking bot crawlers much more aggressively.

Static sites are a lot less impactful, and I usually allow those. I've got a different rule set for them which blocks the known AI scrapers but allows search indexers (though that distinction is slowly disappearing).

And by blocking search robots, do you stop being present in the search results or are you still present, but they do not show the content in question?

I block bots by default, and that prevents them from being indexed since they can't be crawled at all. Searching "dubvee" (my instance name / url) in Google returns no relevant results. I'm okay with that, lol, but some people would be appalled.

However, I can search for things I've posted from my instance if they've federated to another instance that is crawled; the link will just be to the copy on that instance.

For the few static sites I run (mostly local business sites since they'd be on Facebook otherwise), I don't enforce the bot blocking, and Google, etc are able to index them normally.

I wonder if you could have an instance federated to every other instance just for archived purposes, to save the data on every other instance's post and comment. Because copies of posts and comments are saved to federated instances, too, right? Or do I understand the tech wrong?

So it could have an admin team but no users, to prevent people worried about spammers and bots joining that instance to get around defederation rules. Maybe it just has a bot that crawls Lemmy, looking for instances to federate to. Could that work?

Right, but having a centralised search index thingy is better than none at all. Maybe there could be something where it's a joint effort from admins from many of the biggest servers, idk if that would work.

Ah that makes sense not really familiar iwth this stuff so didn't think it's that intensive lol

They usually only index text though

from the top of my head, that won't include lemm.ee, sopuli, beehaw, szmer.info, slrpnk.net, sh.itjust.works, or other threadiverse instances like kbin/mbin.

You'd miss instances that don't use "lemmy" in the URL, but it's at least a better solution than specifying a single instance.

Appending (intext:“modlog” & “instances” & “docs” & “code” & “join lemmy”) to your search query will search most instances. Works with Google, Startpage, SearXNG afaik.

That has not at all been my experience over the years.

Yes, the vast majority of questions are "beginner questions". Which... is true no matter where you go.

But when someone has the ability to articulate a "real" problem? Everyone comes out of the woodwork because that is actually interesting. And there is a very strong communal feeling of "we all have the same problem and are trying to collect data"